Center for Economic Analysis

Using a firm-level data from urban and rural enterprises (n=8,136) provided by USDA's National Survey of Business Competitiveness, we explore factors contributing to the rural-urban firm innovation gap. Our study considers 40 measures of innovation captured by the survey: from initial inputs (such as R&D), to intermediate outputs (such as patents), to final outputs (such as increased market share), and a wide range of other firm characteristics. We use these measures to identify similarities and differences in both innovation and characteristics of rural and urban firms. We conduct our analyses explaining these differences, first by constructing logit models for each category of innovation, controlling for firm age, labor structure, capital structure, industry, and census region, and then employing a Cragg hurdle model to estimate factors associated with number of patents. Based on this analysis, we find significant differences between rural and urban firms in only 6 of the 40 (15%) innovation measures, indicating that firm-level characteristics and/or industry account for most of the observed variation between urban and rural firms. The five most important factors to the rural-urban innovation gap are in order, the portion of the workforce in production, firm age, the portion of the workforce in natural resources-construction-maintenance, whether the firm offers paid maternity leave, whether the firm offers health insurance, and total employment. Firm-level controls account for 92.6% of the rural-urban innovation gap while industry dummies account for only 7.4%. In terms of the number of patents by patenting firms, offering health care coverage was significant and positive, while the urban dummy was not significant at either stage of the Cragg model. One policy implication of our analysis might be that improvements to rural firm innovation may need to be focused more on rural firm characteristics and less on the rates of innovation adoption and creation.